Overland flow measurement

Water moving over slopes towards stream channels is called overland flow, and is generated either by rainfall failing to infiltrate because of a low infiltration capacity (Section 5.1.1) or by precipitation falling on to already saturated ground or the exfiltration of subsurface flow prior to reaching a channel (see Section 1.2) or by fluvial flooding, where the river overtops its banks. Strictly, overland flow is any lateral flow above mineral soil horizons, and thus includes flow within the litter layer. While the proportion of precipitation travelling to channels by overland flow is now considered to be much less than that travelling to channels by subsurface pathways (Section 1.2), overland flow observations are of fundamental importance to the quantification of floodplain inundation (Bates et al., 2006), hillslope erosion and sediment delivery (Owens and Collins, 2006), and the migration of phosphorus across agricultural landscapes (Withers and Bailey, 2003). This brief section will describe methods of measuring the presence and rates of overland flow, and is included to support later modelling (Sections 12.8.4 and 12.8.5) and to assist those hydrologists working on hydro-geomorphological or water-quality problems.

6.5.1 Measurement of incidence and spatial extent

The spatial distribution of areas likely to generate overland flow can be measured using a network of crest stage recorders (Fig. 6.10; Beven and Kirkby, 1979; Burt et al., 1983; Holden and Burt, 2003; Bracken and Kirkby, 2005). These are small cups that fill with water when overland flow is present on slopes. Nests of cups can be used to identify different depths of overland flow.

In areas where overland flow is produced in areas of saturated topsoil rather than areas of low infiltration capacity (see Section 5.1.1), then measurements of moisture content of the organic surface horizons or topsoil (also called A soil horizon) are useful for mapping the likelihood of overland flow (Western and Grayson, 2000; Lin, 2006). Explanation of techniques for measuring soil moisture content is given in Section 6.1.

For erosion studies, knowledge of the micro-scale patterns of overland flow can be important. On soil slopes, overland flow is more likely to be present within micro-rills rather than as thin sheets of water. As erosivity of overland flow is dependent on the depth of flow, concentrating flow into micro-rills makes the water more erosive. Quantification of the three-dimensional structure of these micro-rills can be obtained by surveying with a total station, differential global positioning system (GPS) or laser scanning system (Chihua et al., 2002).

At a much larger scale, remote sensing can be used to identify areas that are covered by overland flow. Airborne synthetic aperture radar (SAR) imagery has been used to map the extent of overland flow caused by fluvial flooding (Bates et al., 2006).

6.5.2 Direct measurement of overland flow

The volume of overland flow generated by a bounded plot can be measured by directing the water from an overland flow trough into either a large tipping-bucket device or a flume. Where a tipping-bucket device is used (Fig. 6.11), the bucket tips when it has filled to a known (calibrated) volume. As it tips, it moves a magnet past a reed switch, which makes an electrical circuit.

The time that this circuit is made can be recorded using a data logger. For details of the alternative flume-based approach, including ways of measuring stage within the flume and calibrating stage to spot discharges, see Chapter 7. The plot is best bounded or isolated with the use of steel or plastic sheets. By bounding the plot,

Fig. 6.10 Minimum depth of overland flow from the peat surface in micro-basin HI on day 239-240, 1999 as monitored by crest-stage tubes: (a) 0300 day 239; (b) 0900 day 239; (c) 2100 day 239; and (d) 0900 day 240. (Reproduced from Holden and Burt, 2003, with permission of the American Geophysical Union.)

Fig. 6.10 Minimum depth of overland flow from the peat surface in micro-basin HI on day 239-240, 1999 as monitored by crest-stage tubes: (a) 0300 day 239; (b) 0900 day 239; (c) 2100 day 239; and (d) 0900 day 240. (Reproduced from Holden and Burt, 2003, with permission of the American Geophysical Union.)

the volume of overland flow per time interval per unit area (hence a depth per time interval, e.g. mmh-1) can be compared with the rainfall depth per time interval or river runoff per time interval. Fig. 6.12 shows a time series of overland flow per unit area recorded by the tipping-bucket shown in Fig. 6.11 (site 'E8') during an example storm-event, together with flows from a road drain ('E2'), a channel-head gully ('E5') and the third-order stream ('P1').

The contact between the soil and the trough that directs the overland flow into the tipping-bucket or flume is important. A large drop between the soil and the trough would produce local erosion of soil immediately upslope, while a step-up can produce artificial ponding and sedimentation (Hudson, 1957). Consequently, the presence of substantial lowering of the soil surface within the bounded plot by erosion would necessitate adjustment of the height of the sill of the trough.

Fig. 6.11 A 3-L tipping-bucket device used for measuring overland flow.

"D

14 15 16 17

Hour on 18 Dec 1995 (5-min sampling)

"D

Fig. 6.12 Overland flow per unit area recorded by the tipping-bucket shown in Fig. 6.11 (site 'E8') during an example storm-event, together with flows from a road drain ('E2'), a channel-head gully ('E5') and the third-order stream ('PI'). (Reproduced from Chappell et al., 2004 with permission of Wiley-Blackwell.)

Overland flow, even within similar topographic locations, has a high spatial variability. Consequently, many bounded plots are needed to determine an accurate catchment estimate of overland flow; care is also needed to avoid biasing the sampling to areas with a high overland flow (Hudson, 1993).

It can be noted that similar troughs inserted into the upslope wall of soil pits (called throughflow troughs) have been used to examine subsurface flow. Knapp (1970) has however shown that the excavation of a pit substantially alters the incidence, magnitude and patterns of subsurface flow, making data from throughflow troughs very difficult to interpret. The exception to this is where throughflow troughs can be inserted in to natural soil faces, such as those studied by Woods and Rowe (1996).

0 0

Post a comment